Despite calculations suggesting that that more than half of the world's population is bilingual (Harris & McGhee-Nelson, 1992), theories of human memory rarely address bilingualism or language proficiency. In fact, at present, existing data are insufficient to allow the incorporation of bilingual proficiency into models of long- term memory. The proposed research focuses on uncovering the mechanisms by which bilingualism and language proficiency impact verbal memory. Verbal memory performance depends in large part upon the strengths of various types of cognitive associations. In the proposed research, a theory is developed based on the idea that the strengths of these different types of associations vary as a function of bilingual proficiency, or for a bilinguals more fluent language (L1) and less fluet language (L2). Each specific aim focuses on the properties of a different type of association. Specific Aim 1 is to test the hypothesis that the ability to form new word-context associations is stronger in L2 than in L1 by testing whether L2 encoding episodes are more distinctive than L1 episodes. Specifically, Experiments 1A and 1B will examine source recognition performance, and Experiment 2 will examine list discrimination performance. If the hypothesis is correct, then bilinguals will exhibit better performance in L2 than in L1 for these tasks. Specific Aim 2 is to tst the hypothesis that the ability to form new temporal associations is diminished in L2 relative to L1 by examining indicators of the temporal associations formed at study. Experiment 3 will examine rehearsal patterns, and Experiment 4 will examine the tendency for items presented consecutively at study to be recalled consecutively at test. If the hypothesis is correct, then L1 words will be rehearsed more often and to a further point in the study sequence than L2 words. Also, L1 words studied consecutively will be more likely to be recalled consecutively than L2 words studied consecutively. Specific Aim 3 is to test the hypothesis that the ability to utilize pe-existing semantic associations is diminished in L2 relative to L1. Experiment 5 will examine the extent to which list organization helps recall performance and the degree of semantic clustering in recall output. If the hypothesis is correct, then organization will help less in L2 than in L1, nd semantic clustering will be weaker for L2 than for L1. The five proposed experiments will be conducted with English-dominant and Spanish-dominant Spanish-English bilinguals and monolingual English-speaking comparison groups. With the specific aims accomplished, we will have a more complete picture of the associative properties that are critical to developing a plausible theory of bilingual proficiency and memory. The information obtained will allow us to further develop a theory based on the idea that language proficiency moderates the strengths of the different types of associations under investigation. This theory will be used in conjunction with other theories of memory to explain a range of bilingual memory phenomena.